I'm planning on building a multi-band vacuum tube receiver. I've got a circuit diagram in hand for a two triode twin reflex (P.E. Edelman, Reflex Radio Receivers, 1924, Fig. 35), and expect to use modernized plug-in coils to cover broadcast AM through at least 40 m (possibly 15 m if I can get it to tune that high). I may add a third tube as a detector in order to use regeneration.
A "reflex" circuit, in this context, makes one or both tubes do double duty; in this case, both tubes act as RF amplifiers (with tuned input and a tuned crystal detector), then the detector output is fed back to the first tube's grid, and from there to the second, with the (intended) result to get a radio with both the high sensitivity and selectivity of a two-staged tuned RF input, and the ability of a two-stage audio amp to directly drive a speaker. Capacitors, chokes, and audio transformers are used to ensure the RF and AF get appropriately routed through the tubes. Double tuning and separate regeneration control promise a "three hands" operation experience, but the final performance makes it worth attempting.
I've got a pair of tubes in hand: 1T4 miniature pentodes. They use relatively low plate voltage (22.5 up to 90 V B+), compared to the older types, but with two amplification stages in RF and two more in AF, I should still get good results. However, as noted, these are pentodes, and the original circuit (which dates from the 1920s) called for triodes. The manufacturer documentation for this tube type calls it an RF "remote shutoff" -- apparently because the screen grid voltage can be manipulated to control gain.
I the suppression grid in this type is internally connected to the filament/cathode, and I've seen a number of examples showing the screen grid connected to the plate via resistance, so that some portion of B+ acts to accelerate electrons after they pass the control grid.
Is there anything else I need to watch out for in converting this kind of reflex/regen receiver from triodes to pentodes?